Log in

View Full Version : Windows Media Player 8.5 and Xscale


Jason Dunn
10-18-2002, 07:58 PM
There have been a lot of questions around whether or not WMP 8.5 that comes with EUU3 is optimized for the Xscale processor. The answer is no. I received the following information from a Microsoft Windows Media Player Program Manager:<br /><br />"One of the key features of Windows Media Player 8.5 for Pocket PC was overall performance improvements which are across the board no matter what processor you are running. These performance improvements are not XScale (in particular Cotulla) specific."<br /><br />So there you have it. Probably not what you wanted to hear, but I'm happy to have performance enhancements that improve my experience on the StrongARM-based XDA. And until you Xscale owners try this player and see what these performance improvements do for you, it's too early to cry foul. :) <!><br /><br />Derek Brown made the following statement regarding this issue <a href="http://discussion.brighthand.com/showthread.php?s=b378f87ba63120559e13cc0df93a4159&threadid=62794">in a Brighthand thread:</a><br /><br />"We made a decision several years ago to move away from supporting several processor architectures and target a single core. We believe this ultimately benefited our OEMs, developers and customers by unifying our platform around a single processor architecture -- ARM V4. <br /><br />The PXA250 utilizes the ARM V5 instruction set with backwards compatibility for ARM V4. When we completed the Pocket PC 2002 software in June 2002, we optimized for the most broadly compatible processor core available at the time (ARM V4), which it still remains today. Choosing to support one processor core ensures we don’t fragment our platform for developers and cause extra work for our ISVs to optimize their applications each time a new processor technology is released.<br /><br />By staying with an ARM V4 architecture we assure longer life for our customers existing hardware – for instance if we were to move to an ARM V5 architecture we would have to obsolete the all SA1110 iPAQ devices.<br /><br />Microsoft considers mobile devices a strategic business. We are committed to working closely with Intel and other silicon vendors on delivering future versions of our Pocket PC and Smartphone devices. We have released specific software modifications to our OEMs that in total are all of the optimizations we believe are possible to maximize PXA250 performance (without causing incompatibilities for our OEMs and developers).<br /><br />So the answer to your question is no unless it could meet the broader requirements listed above. <br /><br />Derek Brown"

Ce
10-18-2002, 08:34 PM
What can we expect from all those upcoming 300 MHz xscales knowing that the current 400 MHz version is only almost as fast as a Strongarm....or significantly slower when using WMP for movies?

pt
10-18-2002, 08:42 PM
here are some pics:
http://www.pocketpcmagic.com/wmp85/

cheers,
pt

JonnoB
10-18-2002, 08:57 PM
What can we expect from all those upcoming 300 MHz xscales knowing that the current 400 MHz version is only almost as fast as a Strongarm....or significantly slower when using WMP for movies?

I thought the same thing. Consumers will be 'upgrading' from 206Mhz to 300Mhz and actually get less with the current trend... except for maybe better battery life.

Charles Pickrell
10-18-2002, 09:09 PM
I understand not wanting to fragment the market, but nearly ALL new Pocket PC devices are utilizing X-Scale now. Since each vendor has to re-release the EUU3 for each model of device, why didn't Microsoft offer them an EUU3 update with WMP 8.5 that had X-Scale optomization and one without. It represents nearly 100% of all new Pocket PC sales.

Sslixtis
10-18-2002, 09:39 PM
Why can't I cry foul til I've tried it? Bloody FOUL on MS for being so slow to support the people who are expanding their market share! They don't seem to have this problem with PCs anymore, although they did at one point.

To me this smacks of the Palm, "I'm the best so I don't have to improve" mentality. If Intel is going to have to design and build the chips then write the software for it what the H*** do we need MS for???? :evil: Are they trying say Intel has better programmers than MS or that they just don't give a damn about the people buying all of the new Xscale Pocket PCs?

I've been upgrading my PPC every 6 months or so for over 2 years now, but if MS wants to start playing games with support I can stop contributing to their bottom line. While I've always been a Microsoftie they are starting to alienate me as a consumer.

My PPC is my friend and as such deserves the best not some lame excuse like "we don't want to fracture the market" (by actually supporting the newer, faster chips). If they felt that way why aren't they still only making DOS for the 8088 PCs? They were more than willing to "fracture" the PC market every time a new chip came out! I guess when it suits their purposes it is ok to "fracture the market", but not for something as lowly and undeserving as supporting the people who are helping them take market share from Palm every quarter.

It used to be the Zen of Palm - "Our users don't want all that stuff in a handheld" Now it's the We Care MS - " We care about all our users with StrongARM so we can't support Xscale" Like they are mutually exclusive! :evil:

Rant over for now :twisted:

Jonathan1
10-18-2002, 09:44 PM
Another slap in the face from MS development. Glad I'm still on a StrongARM based device. :?

Is this the sort of thing that could stall Pocket PC sales this fall?

Dang 3rd post in a row that is hacking on MS today. I must be in a mood. :? Maybe I need to go home and listen to some mood music. ;) Hmm maybe the activation WMA from Windows XP. that's kinda nice. :)

vincentsiaw
10-19-2002, 12:15 AM
dear microsoft, hope you guys listen to our wish an make it come true, if not then we will find a better player hehhehehehe :lol:

farnold
10-19-2002, 08:10 AM
dear microsoft, hope you guys listen to our wish an make it come true, if not then we will find a better player hehhehehehe :lol:

Don't they give us their XScale answer for quite a while sounding like: "Dear customer, get stuffed. You don't have an alternative to our OS anyway. So why should we actually do anything else then telling you how great we and our ridiculous products are. OK, we helped spreading the fairy tale how fast XScale would be while actually making sure it would never happen. If you really dislike it so much... go and get a better alternative... hahahahaha".

I wish SONY lots of success with their devices. Not that I'd ever buy one. But that's the only voice MS actually understands

Jason Dunn
10-19-2002, 06:43 PM
I understand not wanting to fragment the market, but nearly ALL new Pocket PC devices are utilizing X-Scale now. Since each vendor has to re-release the EUU3 for each model of device, why didn't Microsoft offer them an EUU3 update with WMP 8.5 that had X-Scale optomization and one without. It represents nearly 100% of all new Pocket PC sales.

But how do you know that there won't be Pocket PCs released in the next year that will be running on an ARM processor from Motorolla, TI, etc.? If Microsoft optimises for Intel's XScale, what does this mean for devices running ARM processors from other manufacturers? It means that Microsoft has to go back to having multiple versions of their software just like back in the ARM/MIPS/SH3 days. Do you want to go back to that chaos? I sure as heck don't - it was really confusing for the consumer.

Jason Dunn
10-19-2002, 06:53 PM
Why can't I cry foul til I've tried it? Bloody FOUL on MS for being so slow to support the people who are expanding their market share! They don't seem to have this problem with PCs anymore, although they did at one point.

Slow down and think this through for a minute will you? Let's say Microsoft had two choices here (and I believe they did):

1) Make WMP 8.5 optimized for the Xscale processor, giving a 20% performance increase. WMP 8.5 for StrongARM has no performance increase.

2) Make WMP 8.5 optimized for ARM4, which covers StrongARM and Xscale, giving a 20% performance increase.

Which is the better option? #2 of course! That means that EVERYONE, me included with my StrongARM-based XDA, get more performance.

The thing that everyone here seems to be forgetting is that if we're seeing performance DECREASES it's because Intel did a crappy job with their ARM4 implementation. I thought that this was understood, but I see that's not the case. I feel a rant coming on... :twisted:

Ed Hansberry
10-19-2002, 07:41 PM
But how do you know that there won't be Pocket PCs released in the next year that will be running on an ARM processor from Motorolla, TI, etc.? If Microsoft optimises for Intel's XScale, what does this mean for devices running ARM processors from other manufacturers? It means that Microsoft has to go back to having multiple versions of their software just like back in the ARM/MIPS/SH3 days. Do you want to go back to that chaos? I sure as heck don't - it was really confusing for the consumer.

True, but it can be ARM V5 optimized vs ARM V4. I know, that means two optimizations. Either that or we have 400MHz ARM v5 chips running at or below 206MHz StrongARM. :(

Jason Dunn
10-19-2002, 08:28 PM
True, but it can be ARM V5 optimized vs ARM V4. I know, that means two optimizations. Either that or we have 400MHz ARM v5 chips running at or below 206MHz StrongARM. :(

Blame Intel for that - they're the ones that managed to produce an ARM5 chip with sub-standard ARM4 execution. That's just like releasing a 64-bit processor for a 32-bit OS, then complaining about it not being as fast as the native 32-bit processor. Intel did EXACTLY the same thing with their 1st generation 64-bit processors - they were slower at executing 32-bit code than AMD's 64-bit processor, so guess what they did? Back to the drawing board.

Samsung is releasing a 1.2 Ghz ARM processor, and it will be running in their Pocket PC, so it would be a mistake for MS to optimize just for Xscale.

Sslixtis
10-19-2002, 11:13 PM
Jason DunnSamsung is releasing a 1.2 Ghz ARM processor, and it will be running in their Pocket PC, so it would be a mistake for MS to optimize just for Xscale.

True Samsung is making an ARM processor but it is ARMV5 just like the Xscale so who cares if it is 1.2Ghz what's that after the OS downstepping because of lack of ARM V5 Optimization, maybe 500Mhz. And I wouldn't be jumping on that "Intel did a crappy job bandwagon" so fast there either Jason, wait til you see Mototorola and TI's new ARM chips, which are ARMV5, by the way.

I understand you have a vested interest in defending their position at MS and I must admit that I don't really care about an Xscale optimized PPC WMP8.5 :lol: I really don't notice anything wrong with my Xscale running at the equivalent of the StrongARM as far as MP3s go. I don't however, believe it would be that much of a nightmare for them to have released an ARMV4 and ARMV5 version.

If nothing else MS could've said "Yeah, it was more trouble for us to do this and we couldn't do the whole update ARMV4 and ARMV5, but just to show we care about ALL of our customers here is a Media Player for both groups, thanks for your continuing support, MS". At which point right about now we would all be saying "WOW, look at MS aren't they awesome"! It doesn't really take alot to make us happy, just a little acknowledgement of our existence would be nice.

MY main concern however, is Pocket PC 2003, if it isn't ARMV5 Optimized it will not matter how much, how long, or how loud you sing the hymns of MS and the benefits of not "fracturing the market" i'll be done with PPCs until such a time as they return to an OS that is made to fully utilize the CURRENT generation of CPUs at the time.

Janak Parekh
10-20-2002, 12:05 AM
True Samsung is making an ARM processor but it is ARMV5 just like the Xscale so who cares if it is 1.2Ghz what's that after the OS downstepping because of lack of ARM V5 Optimization, maybe 500Mhz.
How do you know the Samsung will have the same efficiency problems executing the ARMv4 instructions as does the Intel XScale?

By the way, Intel has, time and time again, used high clock speeds to mask inefficiencies running old code on the x86 platform--when the 486, the Pentium, the Pentium Pro, and the Itanium came out, they all handled code slower than their previous respective generations. It was the combination of faster clock speeds and later-down-the-road optimized software that took care of this difference.

I don't however, believe it would be that much of a nightmare for them to have released an ARMV4 and ARMV5 version.
MS used to maintain SH3 and MIPS versions of everything for PSPC. Plus ARM on PPC2000. It was quite the nightmare to maintain. There was code everywhere that had one platform and not the others. Even now that's the case for games and such.

At which point right about now we would all be saying "WOW, look at MS aren't they awesome"! It doesn't really take alot to make us happy, just a little acknowledgement of our existence would be nice.
I have a StrongARM-enabled 3870 as my main PPC. To some extent, I agree with you, yet at the same time I'm glad I'm using the same codebase, with the same bugs (or lack of them), as you are -- it simplifies support and compatibility.

MY main concern however, is Pocket PC 2003, if it isn't ARMV5 Optimized it will not matter how much, how long, or how loud you sing the hymns of MS and the benefits of not "fracturing the market" i'll be done with PPCs until such a time as they return to an OS that is made to fully utilize the CURRENT generation of CPUs at the time.
And where are you going to go? According to threads like http://www.spug.net/forums/showthread.php?s=43393dfe6264c28b7376de64d9bd0e86&threadid=32768 (I saw it in http://www.pocketpcthoughts.com/forums/viewtopic.php?t=4137&highlight= ) the ARM-enabled Palms are not perceptibly faster than the Dragonball units until... yes, you guessed it... all their code is optimized. That will not be completely done till OS 6.

--bdj

Jason Dunn
10-20-2002, 12:18 AM
True Samsung is making an ARM processor but it is ARMV5 just like the Xscale so who cares if it is 1.2Ghz what's that after the OS downstepping because of lack of ARM V5 Optimization, maybe 500Mhz.

How the heck would you know? If Samsung did their homework and created their ARM4 execution to be as fast as the StrongARM was, guess what? A 1.2 Ghz Samsung CPU will be much, much faster than 206 mHZ StrongARM. Unless you work for Samsung, it's all pure speculation at this point. But I'm hoping that Samsung will do a good job with ARM4 and my point will be proven. Intel has a long history of doing really poor jobs at optimising their CPUs - they make up for it with sheer clock speed. Do some homework and you'd realize this is no different than what happened with the Pentium Pro, P4, etc.

And I wouldn't be jumping on that "Intel did a crappy job bandwagon" so fast there either Jason, wait til you see Mototorola and TI's new ARM chips, which are ARMV5, by the way.

So what if they're ARM5 - the question is, how do they handle ARM4? From what I've been told by developers, ARM5 isn't a drasticly new set of instructions - it's more like a simple add-on for ARM4. ARM4 is where we're at now with the Pocket PC OS, so ARM4 execution speed is what matters. Compared to the huge number of StrongARM devices out there on the market, the newer Xscale devices are just a drop in the bucket - it would be foolish for MS to support ARM5 at the exclusion of ARM4.

I understand you have a vested interest in defending their position at MS

No, actually I don't, and I deeply resent you and your implication that I do. I call MS on bad moves where I see them, and I defend them where mob hysteria runs afoul of common sense. This is one such case.

I don't however, believe it would be that much of a nightmare for them to have released an ARMV4 and ARMV5 version.

I humbly submit that you have no clue what you're talking about.

Mojo Jojo
10-20-2002, 01:37 PM
Just something quick to toss out...

How does it make it easier for the 'average consumer' to have only one set of code to worry about, versus an 'average consumer' having to know which processors and processor speeds work best with an OS?

I submit to you that in reality your trading one consumer unfriendly scenerio for another. I think readers here take for granted the knowledge that is gathered by the collection of this and other sites regarding what works best etc...


--end of main rant, read further for complete rant or end reading now --

It is my opinion that for PDA's to reach mainstream people need to stop making decisions for them. Do not limit what MY PPC can do because it may confuse Him or Her . Instead how about creating software that is more robust? Back in the early days there was chaos in programs because they were created by programers with limited resources, they may have programmed for MIPs only because that is what they had. Now I ask you... who was responsible for the chaos? The OS... or the Programmers? I think people unfairly blammed the OS because small firms/programmers/companies didn't have the ablity to program for all models...

After that Microsoft decided, not you or I, that to fix things they had to limit things for the brand. Essentially what you have happening is Technology Censorship instead of an open enviroment. Much like trying to read 'Catcher in the Rye' if your town or school band the book... do you blame the author of the book or the town/school?

While the corrolation between the two are less defined what you have is someone making a decision about how well your PDA works, instead of letting you make the choice.

Now in the case of the media player this isn't such a problem, I have already installed multple players that work better then the window's player. However I can not go out and find an entire OS clone that runs better and remove the MS one.

Some might say I can go by a Palm or Linux model if I do not like this. Normally I would agree, but my work enviroment and the tasks I need to do limit my choices to only one, Microsoft.

Now before you hit the reply button and come at me guns blazing that you think I am demanding MS to do something different, I am not. I am however suggesting that if they took another approach they could please more users, and in turn do better business.

In business being the leader in a field requires that the group take on more responsiblites then their rivals, least they be replaced by someone willing to do so. To knowingly sit back and not do something simply because others can not do so, will not win the loyality of users, it will not increase brand awareness.

To bring this back full circle, when I hear things like, Microsoft is trying to, is limiting to, etc... what I hear is that Microsoft is only willing to do the least amount of things necessary to get a task done because no one else can do it. To me, that means what ever I require from a Microsoft product needs to be functional at time of purchase, I have no belief that Microsoft will strive to beat my needs and try to retain me as a loyal customer now and in the future. As such the moment another product surpasses Microsoft's product I will no longer be doing business with Microsoft. To me... this cause and effect business model is very poor, and moving forward I can fully appreciate why Microsoft is having a tough time breaking into the Corporate enviroment as well as dethroning Palm in sales.

Now if you take these same prinicples and apply it to the Palm brand... you can see where Palm is losing ground to Sony. Palm choose to direct a market by choosing for the customer what they did and did not need in a PDA, all those that followed suit (Handspring / Visor) did not succeed in market there because they only played runner up, the moment a company surpassed that model (Sony) did the consumer vote with their dollars and start switching.

In the Microsoft case, there simply isn't a competitor (yet) leaving a begruding choice to users, Microsoft or nothing. That leaves the PPC makers in a rough place as they can not innovate the product with out Microsoft support. When trying to capture consumers that leaves only one avenue left, price. In the upcoming price wars from Dell and Viewsonic the PPC market will have a low profit model which may make some makers withdraw from the market. Leaving the PPC camp in an awkward position, no longer able to distinguish models by performance means due to Microsoft leveling the field and not optimizing code, due to the price wars soon to take place, what will be left is a PDA made by the lowest bidder with the cheapest parts and quality. Not exactly the traits I look for in my buying decision.

To me the flaw here is more fundemental. In that the software is driving the development of hardware. That the hardware completely caters to the software is backwards from the rest of the computer technoloogy field. Why can we not have a PDA standard (CPU, VIDEO, Storage, Memory) then choose what OS it runs with? Why is it the software dictates my processor? Why do I get penelized for an Xscale versus a StrongARM? My computer has an AMD processor... yet I can choose a Linux or MS product to run on it, we would all be up in arms if MS said, "Well we are going to optimize our code to only run on Intel platforms, we won't make any changes to get it to run on an AMD, but if it does... great. It is up to AMD to create a chip that runs our os."

Do you buy a car that can only run Mobil 92 octane? Not Exxon 92 octane?

Now, that over simplifies a lot of things, I freely admit that. What it comes down to for me is that if Microsoft is the driving force of this technology, it can not, should not, limit itself to the lowest common denominator, least we get PDA's that are the lowest common denominator.

Wow... do I get a prize for the longest quick question catagory? Anways, I just want better code for the product I purchased.

Sslixtis
10-20-2002, 03:08 PM
Thanks, Jason I really like how both you and BigDaddyJ used judicious quoting to make it appear that I was saying that it wouldn't be that much of a problem doing two versions of the OS when I was talking about just WMP8.5. That is kinda low. And I submit, beneath you.

I agree Samsung, Motorola and TI MAY do better with ARMV4 integration than Intel did, however; do you really see OEMs going back to ARMV4 or do you see them all going to ARMV5? Honestly please, is the future V4 or V5? And I did say ARMV5 Optimization not just Xscale after your first reply decrying Intel as a shoddy second rate company. That apparently is just a drain on the computer industry and has been carried by software since the 8088.

And as for the vested interest part, been to any MS functions lately? Somehow I doubt MS invites people to Redmond that are really critical of their products. That is not an attack, It's understandable to defend people you know and companies you like.

ARMV4 is where we were besides the PPC Phone Editions all the new PPCs seem to be ARMV5 to me
iPaq, Dell, ViewSonic, Toshiba, heck even the new Palms are sporting them albeit the low-end ones.

How about we just agree to disagree. The only way I'll believe that Samsung, TI, or Motorola is going to have the UberARM that needs no optimization is to see it. And when, exactly, are they releasing this processor? They had announced they would be mass producing 500-700Mhz ARMV5 processors by the end of July, I still don't see them
much less a 1.2Ghz one. But hey, I wish them the best, if they do come out with a 1.2Ghz UberARM in the not too distant future I will be extremely happy! :lol: Then I wouldn't be writing these bloody posts!

Janak Parekh
10-20-2002, 04:52 PM
however; do you really see OEMs going back to ARMV4 or do you see them all going to ARMV5? Honestly please, is the future V4 or V5?
Both. Smartphone 2002 devices are coming out right now with ARM4. One PPC Phone Edition is using a 133MHz TI OMAP ARM4 processor (the Jornada 928), and another one is coming out. Why? Extremely low cost and decent performance.

ARM4 is far from dead, fortunately or unfortunately. To some extent, it's like asking if the P4 is dead when the Itanium comes out. Or, a better analogy is asking if the Celeron was dead when the P4 came out. Microprocessor manufacturers keep long lifetimes for processors for precisely this reason.

Yes, in 3-5 years we'll see a drastic drop in ARM4 adoption. Not now.

--bdj

Janak Parekh
10-20-2002, 04:58 PM
How does it make it easier for the 'average consumer' to have only one set of code to worry about, versus an 'average consumer' having to know which processors and processor speeds work best with an OS?
For what it's worth, I was one of the first iPaq adopters, way back in June, 2000 when they were the first units with any ARM technology at all. And it was massively frustrating that I couldn't use 90% of the software that was out there. I was reduced to playing the built-in MS games, using Pocket Outlook, and that was it. I was far from the only one.

Even now there are cool apps that I'm about to download, and then I notice it's PSPC and doesn't support ARM. If I had a Jornada 548 they'd run fine.

Now, the tables have been turned around. Quite a few PPC2000 owners wish they could still use their Casio E-125 to run new software, as they honestly believe that was the best PPC ever produced (and it certainly had the best screen, although the new transflectives are comparable to some extent). They can't, though, since developers only release ARM.

In other words - it makes it easier for the average consumer because when they see a "Pocket PC 2002" program, they know it'll work on their device. Whether they have a newer one or an older one. More importantly, it makes it easier for the developers to provide the consumer with a product that works.

One last example: during the PPC2000 days, when new games came out, they tended to only support one device. Why? The author didn't own multiple devices, and despite cross-compilation there's no guarantee a program written on one device will work perfectly on another when recompiled. Often, it took weeks or months for a product to come out for the other devices. (Admittedly, this is not completely solved yet, but most PPC2002 games today have far more cross-product-line support than back in the PPC2000 days.)

This, by the way, is why we're not all using Itaniums and Windows XP 64-bit edition right now. If you go buy one of those boxes, you'll notice nothing is supported natively on it, except Windows XP, and maybe Office, and everything else has to run through x86 emulation (and it's dog-slow, too, much worse than ARM4/ARM5). Again, it's a deliberate decision to keep the convience of the mass consumer in mind.

Also note that we're not the average consumer :)

--bdj

p.s. The Linux model has many problems supporting binary, commercial software precisely because of all the ports of Linux. Most commercial Linux-supported software requires a x86 running RedHat. While multiplicity is a good thing sometimes, consumer convenience is rarely maintained IMHO.

gateway
10-20-2002, 07:00 PM
I know this is mute cause I got my answer from M$ a while back but Im wondering if they snuck in support for the new windows media player 9 codec ?

I heard that it wouldnt be supported till 2003. ? :/

also about the Xscale situation, im pretty unhappy about it, i paid the hefty price for getting the new ipaq with Xscale tech, and come to find out that it was alot slower on many of the apps than my older ipaq. This doesnt sit well with me and im sure there are alot of pissssed off people about the same situation. Im currently using Pocket Divx for most if not all video i play/encode with the vp3 codec it has some pretty amazing quality.

Gateway

mookie123
10-20-2002, 09:07 PM
What is with Microsoft and not supporting Xscale anyway?

ARE they or are they NOT committed to this CPU in pocket PC?

what is with insisting they are maintaining ARM V4 for the sake of smartphone. (see brighthand thread) who cares what they gonna do with smartphone. Smartphone is NOT pocket PC. and most if not all PPC software will not run on smartphone. Is microsoft so greedy trying to grab, cutting corner and emphasize market that doesn't exist yet over the existing pocket PC?

also, the old iPAQ is dead. screw them. update the new OS for Xscale and make a software patch for the old Strong arm iPAQ. (just like MMX and non MMX case) I am sure the old iPAQ owner will tolarate the fact that software designed for Xscale will still run albeit slower.

and more importantly.....Whack the media player division. WTF? What are they getting paid for if not able to create player version for Xscale optimize and another for non Xscale. This media player issues doesn't even jibe with their reasoning about trying to maintain as many similarity between PPC and smartphone. They are trying to imply that one day apps developer will come up with Xscale optimized titles, but they themselves doesn't even do it with media players???

somebody really ought to break up that company to prevent this sort of dubious market move and cost saving! (say, isn't leveraging a product in one market to enter another sort of monopolistic behavior? )

PS. I wonder if this is about the time to do a grass root campaign to force Microsoft to provide Xscale optimized solution instead of keep accepting their lame excuses.

jefito
10-21-2002, 03:19 AM
What is with Microsoft and not supporting Xscale anyway?

They are. PPC2002 works on XScale. Which part of "support" do you not understand?


ARE they or are they NOT committed to this CPU in pocket PC?

They are committed to the ARM4 instruction set, which XScale provides.


what is with insisting they are maintaining ARM V4 for the sake of smartphone. (see brighthand thread)

They are not insisting that, at least with respect to PPCs. Their stance on ARM4 is for the PPC platform, and evidently also the same for Smartphone. Go back and re-read the brighthand thread. Derek Brown makes it pretty clear what their strategy is, and why.


who cares what they gonna do with smartphone. Smartphone is NOT pocket PC. and most if not all PPC software will not run on smartphone. Is microsoft so greedy trying to grab, cutting corner and emphasize market that doesn't exist yet over the existing pocket PC?

It's nothing to do with SmartPhone -- see above.


also, the old iPAQ is dead. screw them. update the new OS for Xscale and make a software patch for the old Strong arm iPAQ. (just like MMX and non MMX case) I am sure the old iPAQ owner will tolarate the fact that software designed for Xscale will still run albeit slower.

As he said, Derek Brown expects that we will be seeing StrongARM PPC designs coming out in the future. Maybe you know something he doesn't?


and more importantly.....Whack the media player division. WTF? What are they getting paid for if not able to create player version for Xscale optimize and another for non Xscale. This media player issues doesn't even jibe with their reasoning about trying to maintain as many similarity between PPC and smartphone. They are trying to imply that one day apps developer will come up with Xscale optimized titles, but they themselves doesn't even do it with media players???

Ummm, the media player division is carrying out MS's stated strategy. Why would you whack them? And stop rattling on about SmartPhone -- it's not the issue.


somebody really ought to break up that company to prevent this sort of dubious market move and cost saving! (say, isn't leveraging a product in one market to enter another sort of monopolistic behavior? )

?????


PS. I wonder if this is about the time to do a grass root campaign to force Microsoft to provide Xscale optimized solution instead of keep accepting their lame excuses.

Go for it. But given Derek's explanation of MS's strategy, it seems unlikely that you would be able to "force" Microsoft to do anything.

Jason Dunn
10-21-2002, 05:55 AM
I know this is mute cause I got my answer from M$ a while back but Im wondering if they snuck in support for the new windows media player 9 codec ?

No, 9 Series support won't be coming until later on (probably not this year). Bummer, because I really want to use it, but I guess that codecs haven't been ported to CE yet - which I think means they weren't working in tandem. Codecs get finished for X86, THEN they get handed off to CE guys for porting.

Jason Dunn
10-21-2002, 06:00 AM
How does it make it easier for the 'average consumer' to have only one set of code to worry about, versus an 'average consumer' having to know which processors and processor speeds work best with an OS?

If you had a Pocket PC in early 2000 when there were three CPUs, you'd understand how silly your statement was - it was CHAOS when a consumer went to a vendor web site, wanted to download a trial application, and was then presented with MIPS, ARM, SH3. They know they bought a Casio E-125, but what the heck are these strange words? Pain my friend, consumer pain. Forcing everyone to go to ARM was a very, very smart move by Microsoft.

And here's a big reality check for you and the others who are getting so upset about this: the average consumer doesn't care. Period. Pick up a StrongARM or Xscale Pocket PC, and use it - it will be fast. Joe Consumer isn't going to benchmark video FPS and compare them - he'll use it, and if it's snappy and responsive, he'll be happy.

Power users live on a different planet than normal people, and sometimes it drives me crazy when they don't realize or understand that - a lot of the issues that we rant about on this site are applicable to perhaps 1% of the buying public. We have to remember that!

Sslixtis
10-21-2002, 06:58 PM
Jason Dunnthe average consumer doesn't care. Period. Pick up a StrongARM or Xscale Pocket PC, and use it - it will be fast. Joe Consumer isn't going to benchmark video FPS and compare them - he'll use it, and if it's snappy and responsive, he'll be happy.

True, all true. I guess you could say I'm in that Power User group, and for most things on the PPC who cares about optimization for V5.

Video playback however, is one point where it is really noticable. Heck, I don't care if they charge more for an optimized player, I just want the option.

Have you tried the PocketTV Enterprise Edition on an Xscale? All I can say is WOW! But what about all those WMVs I have, they still play like crap on my PDA :cry: (Well comparatively speaking anyway) If MS isn't willing to support V5, couldn't they atleast let 3rd party software companies supply WMA/V support? Does anyone know if MS has offered anyone that option?

mookie123
10-21-2002, 07:10 PM
They are. PPC2002 works on XScale. Which part of "support" do you not understand?

V4 is a subset of what Xscale can do. IE. PPC2k2 is not fully utilizing Xscale which is the crux of the issues. This move can be seen as a hedge supporting other V4 capable chips from TI and Moto. (hence maybe they are not commited)

all future models of ARM chips for PDA will support some sort of extension which is incompatible with the other. V4 is the only common dinominator between Omap, MX, Xscale.

>>what is with insisting they are maintaining ARM V4 for the sake of smartphone. (see brighthand thread)

They are not insisting that, at least with respect to PPCs. Their stance on ARM4 is for the PPC platform, and evidently also the same for Smartphone. Go back and re-read the brighthand thread. Derek Brown makes it pretty clear what their strategy is, and why.

To Quote:

"Microsoft considers mobile devices a strategic business. We are committed to working closely with Intel and other silicon vendors on delivering future versions of our Pocket PC and Smartphone devices. We have released specific software modifications to our OEMs that in total are all of the optimizations we believe are possible to maximize PXA250 performance (without causing incompatibilities for our OEMs and developers)."

"It is important to remember that xScale is Intel's ARM product. Our planning assumption needs to be that there will be Smartphones and Pocket PCs on non-Intel ARM processors. And it is important to remember that while we won't have binary compatibility for applications between Pocket PC and Smartphone that we do strive to have as much shared code as possible between the products which includes compatibility down to the ARM level. "


http://discussion.brighthand.com/showthread.php?s=5654f1ab14c1563d4ed7099d9ceb7eb7&threadid=62794&perpage=15&pagenumber=1


It's nothing to do with SmartPhone -- see above.

not according to Microsoft apparently.


As he said, Derek Brown expects that we will be seeing StrongARM PPC designs coming out in the future. Maybe you know something he doesn't?

Do YOU know someting that we don't? since Derek didn't say that. The only strong arm product actively being pushed right now is XDA.


Ummm, the media player division is carrying out MS's stated strategy. Why would you whack them?

What is MS stated strategy? screwing the customer by just developing the least posible number of software with the most common dinominator? The market fragmentation reasoning doesn't jibe in media player case. since it IS an app instead of OS. It cannot run on other class beside PPC and can enhance performance greatly by using some of Xscale extentions.

Go for it. But given Derek's explanation of MS's strategy, it seems unlikely that you would be able to "force" Microsoft to do anything.

do you actually BELIEVE that goobly gook given the inconsistencies??
amazing.

Jason Dunn
10-21-2002, 07:58 PM
But what about all those WMVs I have, they still play like crap on my PDA :cry: (Well comparatively speaking anyway) If MS isn't willing to support V5, couldn't they atleast let 3rd party software companies supply WMA/V support?

Tell me what you think of the performance once you install EUU3 and get WMP 8.5 - I've seen a significant improvement on my XDA, but it's hard to measure because I didn't have any prior benchmarks.

I completely agree that performance is important, but if Microsoft made 8.5 work faster on BOTH StrongARM and Xscale, that's good right? We all win. Who cares whether it's ARM5 or improved ARM4 code? Better is better. :D

Mojo Jojo
10-21-2002, 08:13 PM
If you had a Pocket PC in early 2000 when there were three CPUs, you'd understand how silly your statement was - it was CHAOS when a consumer went to a vendor web site, wanted to download a trial application, and was then presented with MIPS, ARM, SH3.

Down boy... I agree that in Pocket PC 2000 days it was fustrating. What I was trying to say was would it have been less fustrating if the developers had all three versions instead of just one? If all three versions of the code was in one downloadable and would select the correct installation for you... wouldn't that have reduced the chaos? I am trying to point out that the choas was chaos for other reasons then just the three processor types. The market was very fan/small developers driven... do you think if oracle/sun/IBM/Adobe/Any large company would only produce code that worked on one version only? However Joe Smith in his basement might be able to only do one version....

As for the rest of my thought... it appears that your impling that for an avergae consumer finding out how well various apps run on diffent processors is easy?!? You walk into any computer store and ask an average user what they expect will happen when running a program on a 206 mhz processor vs an 400mhz processor... and the person will tell you they expect better results on the 400.... Right or wrong that is what you will find. Going with that thought you don't expect the user to feel 'cheated' when his experience tells him differently?

Either way you look at it, its something that has to be learned outside of brochures and company propoganda. Till PDA's can be bought like televisions your going to have issues. Subscribing to a limited mindset regarding differrent codes does not make an effort to solve those issues.

And here's a big reality check for you and the others who are getting so upset about this: the average consumer doesn't care. Period. Pick up a StrongARM or Xscale Pocket PC, and use it - it will be fast. Joe Consumer isn't going to benchmark video FPS and compare them - he'll use it, and if it's snappy and responsive, he'll be happy.

The part I agree here but view the outcome differently is snappy and responsive... those are subjective terms and everyone is differrent. I also think that the theory proposed is off because with out optimization the experience isn't snappy or responsive. The new models do not preform better then a model released a year ago...

And last... a few are getting upset but some of us are trying to have a discussion on the forum board. Do I demand Microsoft do something different... nope, stated that already, I am just suggesting that a different method and approach would reach a larger audience. Heck if you believe I am wrong look at the reaction form the audience now, if the current approach was correct there wouldn't BE so much disagreement.

Jason Dunn
10-21-2002, 08:23 PM
As for the rest of my thought... it appears that your impling that for an avergae consumer finding out how well various apps run on diffent processors is easy?!? You walk into any computer store and ask an average user what they expect will happen when running a program on a 206 mhz processor vs an 400mhz processor... and the person will tell you they expect better results on the 400

No, that's just it - the comparisons won't be done by people. They'll pick up the device, launch the Contacts app, and if it "feels" fast, it IS fast. If I were testing a device, I'd probably load up a video file and see what the performance was like - I did that at CeBIT with the Loox. But my point is that you and I are not normal consumers, so while you rail at the heavens for not having an Xscale optimized OS, the rest of the world hasn't even noticed. The reason? Day to day, the Xscale devices performn just as fast at the StrongARM processors - it's not like contacts takes 200% longer to launch on an Xscale processor.

Yes, video is one area where this is different, but it's not simply a matter of the CPU - it's the slow system bus, it's unoptimized display drivers, etc. Reality check: how many Pocket PC users use video on a day to day basis? Sure, it's a great feature to have and I always carry a video clip or two with me, but it's certainly not the most-used feature. Hence, all the furor over this is a little melodramatic for my tastes.

Heck if you believe I am wrong look at the reaction form the audience now, if the current approach was correct there wouldn't BE so much disagreement.

Someone once told me "A person is smart. People are stupid."

One on one, if you were to call Derek Brown at Microsoft and ask him to explain why they did this, I know that you'd leave that conversation saying "Ok, I don't like the result, but I can see why this was the only logical move for them to make."

On a discussion forum, mob mentality rules - most people are selfish, they care about themselves, and THEY want an Xscale version of WMP 8.5 and THEY want it NOW...never mind the fact that 8.5 performs much better with video and it may make EVERYONE happy, both StrongARM and Xscale users. Nope, that's too logical for the mob, and they'll keep ranting...

Janak Parekh
10-21-2002, 08:33 PM
Down boy... I agree that in Pocket PC 2000 days it was fustrating. What I was trying to say was would it have been less fustrating if the developers had all three versions instead of just one? If all three versions of the code was in one downloadable and would select the correct installation for you... wouldn't that have reduced the chaos?
No. ARM came out after MIPS and SH3. Developers took a lot of time back in 2000 to release ARM support, and it included a raft of testing on the new devices.

PPC2000 did support combined installers. More likely, you'd hit the installer and it would say "platform not supported".

--bdj

Jason Dunn
10-21-2002, 08:36 PM
More likely, you'd hit the installer and it would say "platform not supported".

And what's worse is that until the iPAQ started to gain some serious traction (early 2001), most developers wouldn't release ARM versions because they were unable to get a test device.

mookie123
10-21-2002, 09:08 PM
Why is this thing happening again? this is exactly like the MMX issue, where AMD, Microsfot, and Intel bicker at each other.

WHY can't they learn the lesson and implement a strategy that works with MMX case?

They sure pissed a lot of people off who just spends $5-700 dollar on gadget that didn't perform up to its potential.

And dammit, optimize the media player for each chips! So what if there is a big difference. ie. Strong arm spews 14fps, and Xscale 45fps. that's not going to "fragment the market" That's totally different issue than OS optimization.

grrr....!

Janak Parekh
10-22-2002, 12:50 AM
Why is this thing happening again? this is exactly like the MMX issue, where AMD, Microsfot, and Intel bicker at each other.
Welcome to corporate competition, where vendors are not required to cooperate. ARM is a bit weird in the sense a third party defines the language, but licensees are welcome to come up with their own implementations and optimizations, which is what happened to XScale.

If you think MMX is bad, play around with the DVD industry...

WHY can't they learn the lesson and implement a strategy that works with MMX case?
Neither AMD nor Intel want to pay each other's licensing fees (which are presumably substantial to stifle competition).

They sure pissed a lot of people off who just spends $5-700 dollar on gadget that didn't perform up to its potential.
Are you referring to MMX? If so, the average consumer hardly knows. "Ooh, MMX, 3DNow, cool, it's faster!" Believe it or not, similar things happen with things like PocketPC.

And dammit, optimize the media player for each chips! So what if there is a big difference. ie. Strong arm spews 14fps, and Xscale 45fps. that's not going to "fragment the market" That's totally different issue than OS optimization.
What? Think of the scenario where manufacturers start releasing XScale-only or XScale-specialized implementations. I can see this happening for games, for example. As a 3870 owner, I'm against that for now. A lot of others are too. There's your market fragmentation right there. A majority of PocketPC owners are still StrongARM users. It's like saying that UT2003 basically needs a Pentium IV to be playable... that's suicide for game manufacturers.

You assume that software manufacturers will do the good thing and optimize their code for each processor. Don't forget that costs time and money, and a lot of them don't. Witness the MMX case :)

One other thing, and I think Jason has been trying to say this: you assume that "XScale optimization" will fix everything. How do you know? Whenever a first iteration of a processor comes out, it tends to be inefficient. Increasing clock rate requires deeper pipelines and also places other stringent requirements; XScale apparently has slower memory lines than StrongARM. Generally, cranking up the clock rate even higher overcomes these kinds of problems, which XScale hasn't done yet.

We don't know what performance improvements will come from ARM5. I would indeed like to see that documented by Intel or Microsoft. If there's only, say, a 5% performance increase, then it's not worth it until XScale matures. If there's a 200% increase (as 15 vs 45 fps would indicate), then there's a solid motivation for doing it. However, we don't know. (Actually, it's very likely that you'll get less than 100% - doubling the clock speed rarely causes a double in the actual performance of a processor, even with a new instruction superset...)

--bdj

Janak Parekh
10-22-2002, 12:58 AM
And what's worse is that until the iPAQ started to gain some serious traction (early 2001), most developers wouldn't release ARM versions because they were unable to get a test device.
Believe me, I know. I still have no idea how I managed to get one that June. I placed an order from Insight, and it shipped the next day.

We were also forced, back then, to use HPC ARM programs with the messed-up dialog widths :D

--bdj